Research

Define a target audience and validation hypotheses

Build an evidence-bounded audience definition and turn uncertainty into testable questions

9 min setupTested with:ChatGPTReviewed: 2026-08-28

Add your context

Your text stays in this browser. AILesson Prompts does not send it to a model or server.

Business decisionAvailable evidenceValidation capacity

Your prompt

Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt

Define a target audience without turning assumptions into demographic stereotypes.

Product, market, decision, timeframe, constraints, and success:
[decision]

Research, behavior data, sales notes, exclusions, samples, and contradictions:
[research]

Methods, recruitment, budget, time, sample limits, ethics, and thresholds:
[capacity]

Start with the decision the audience definition must support. Inventory evidence with source, sample, date, observed behavior or quote, and limitation. Define candidate segments primarily by situation, job, trigger, current alternative, desired progress, constraints, and observable eligibility; use demographic attributes only when relevant and evidenced. State explicit exclusions and avoid claims about personality or motivation that were not observed. For each segment separate known, inferred, and unknown. Rank hypotheses by decision impact and uncertainty. Express each as a falsifiable statement with evidence for and against, a low-cost test, sample, measure, threshold, and consequence of passing, failing, or remaining inconclusive. Note recruitment and survivorship bias. End with a provisional audience definition that must not be presented as validated.

Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.

From input to outcome

A worked example

See how concrete context turns this recipe into a usable result

Actual input

Business decision: Choose the first UK segment for a lightweight invoice-reminder tool sold at GBP 12/month. Decide within 3 weeks whether to focus the launch page and onboarding on solo service businesses. Success means a segment we can identify, reach, and observe completing three reminders in 30 days.
Available evidence: Eight interviews recruited from a freelancer newsletter: six manually chase overdue invoices; four say awkward wording delays follow-up; two use full accounting suites and see no need. Product beta: 31 users, 18 sent a reminder, 9 sent three; occupation is missing. Five sales notes from small agencies mention approval workflows, which the product lacks. Interviews occurred this month; beta recruited from existing newsletter readers.
Validation capacity: Can run 10 more interviews and a 100-visitor message test; GBP 500; three weeks. Recruitment through freelancer groups and two accountant referrals. No scraping or sensitive profiling. Proceed threshold must be defined before testing.

Example output

Decision: determine whether solo UK service businesses should be the first launch segment, not whether every freelancer needs the tool.

Evidence map: six of eight newsletter-recruited interviewees manually chase invoices; four described wording friction. This is recent qualitative evidence, but the convenience sample overrepresents newsletter readers. Two interviewees using accounting suites reported no need, supporting an exclusion hypothesis. Beta behavior shows 9 of 31 reached three reminders, but missing occupation means it cannot validate the proposed segment. Agency approval needs are contrary evidence for serving agencies with the current product.

Provisional segment: UK solo service providers who issue invoices, currently follow up manually, lack or avoid an automated reminder in an existing suite, and have at least one overdue invoice trigger. Exclude teams requiring approval workflows and people satisfied with existing automation. Business type, trigger, and current process must be verified during qualification; do not infer discomfort or price sensitivity.

Highest-impact hypothesis: among qualified solo providers, wording or follow-up effort is frequent enough that at least 6 of 10 can describe a recent manual chase and at least 4 will complete a reminder setup using their own scenario. Test with ten interviews plus a task prototype. If both thresholds pass, proceed to the message test; if either fails, revisit problem or segment; if recruitment yields fewer than eight qualified participants, mark inconclusive.

Message hypothesis: problem-led copy about timely, professional reminders produces at least 12 of 100 qualified-page visitors starting setup, compared with a predeclared baseline or control that is currently missing. Do not interpret an unqualified 100-visitor convenience sample as segment demand.

Open decisions: define the message-test control and allocation, track occupation and current tool with consent, specify what counts as completed setup, and ensure referrals do not dominate the interview sample. This audience is provisional until the tests meet their thresholds.

Why this works

1
Behavioral eligibility produces more actionable segments than unsupported personas.
2
Pass, fail, and inconclusive rules make research change a decision instead of merely adding notes.